Lisa Su’s Signature! GMKtec EVO-X2/X3 Shine at AMD AI DevDay
Shanghai, May 19, 2026 — GMKtec, a global leader in mini PCs, made a strong showing at the 2026 AMD AI Developer Day, featuring its full lineup of flagship desktop AI supercomputers, the EVO-X2 and EVO-X3. The products earned personal signature recognition and ecosystem support from Dr. Lisa Su, Chairman and Chief Executive Officer of AMD. GMKtec will deepen its integration with AMD’s ROCm open-source ecosystem to help Chinese AI developers unlock new possibilities for edge-side local large model computing power and jointly build an open, innovative AI hardware ecosystem.
Centered on the core principles of Openness, Building, Innovation, and Connection, this year’s AMD AI Developer Day brought together global AI engineers, technical leaders, open-source developers, and industry ecosystem partners. Focused on cutting-edge areas like edge AI, large model inference, heterogeneous computing, and intelligent agent applications, the event stands as one of China’s most influential AI developer gatherings and a key platform for technical exchange and ecosystem collaboration. As a long-standing, deep strategic partner of AMD, GMKtec leveraged its first-mover advantage and technical expertise in the desktop AI supercomputing space to become a core hardware showcase brand at the conference, offering high-performance mini computing solutions tailored specifically for AI developers.
A standout at the event, the EVO-X2 is the world’s first desktop AI supercomputer powered by the AMD Ryzen AI Max+ 395 processor. Built on a Zen5 architecture with 16 cores and 32 threads, it integrates an XDNA2 NPU and RDNA3.5 iGPU, delivering a total heterogeneous computing power of 126TOPS across CPU, GPU, and NPU. It seamlessly supports local deployment of ultra-large 235B-parameter models, enabling users to run mainstream open-source large models and intelligent agent applications right out of the box. Balancing robust computing performance, low power consumption, and data privacy, it has become the go-to hardware for developers engaged in local AI research, model debugging, and inference training.
The newly upgraded EVO-X3 Desktop AI Supercomputing Center boasts comprehensive improvements in computing power, expandability, and ecosystem compatibility. It features enhanced support for AMD’s ROCm software stack, boosting performance for multi-model parallel inference, distributed training, and long-text intelligent agent operations. Equipped with WiFi7, high-speed USB4.0 ports, dual PCIe4.0 drive bays for massive storage expansion, and an OCulink interface for external discrete graphics cards, it pushes the boundaries of edge AI computing power. This makes it ideal for professional developers, AI studios, and industry clients with advanced computing needs, covering use cases from individual development to enterprise-level applications.
Previously, the EVO-X2 took the stage during Dr. Lisa Su’s keynote at the 2026 CES International Consumer Electronics Show, establishing itself as a benchmark hardware product in AMD’s Ryzen AI Halo desktop AI ecosystem. At the May 19 Developer Day, Dr. Lisa Su personally signed the products, adding significant credibility to GMKtec’s ecosystem partnership. This milestone underscores the deepening collaboration between the two companies in desktop edge AI, the ROCm open-source ecosystem, and developer empowerment.
A GMKtec representative commented: “Leveraging AMD’s leading heterogeneous computing platform, GMKtec remains focused on the desktop AI supercomputing space, dedicated to making high-performance, low-threshold, easy-to-deploy local AI computing power accessible to every developer. Our participation in this AMD AI Developer Day not only brings our ecosystem collaboration to life but also aims to use the EVO-X2 and EVO-X3 as a hardware foundation to help Chinese developers accelerate AI technology innovation and application deployment—together building a more open and inclusive edge AI ecosystem.”
Moving forward, GMKtec will continue to collaborate closely with AMD, leveraging the ROCm open-source platform to partner with domestic open-source communities and technical developers. The company plans to launch more AI hardware products optimized for large models, intelligent agents, and edge inference, covering diverse scenarios such as individual developers, small and medium-sized enterprises, and custom industry solutions. GMKtec is poised to maintain its leadership in the desktop AI supercomputing space and empower the intelligent transformation of industries worldwide.
burningbecks
Brad Linder
Pierre Lecourt
